WOOD SPECIFIC GRAVITY OF TREES AND FOREST TYPES IN THE SOUTHERN PERUVIAN AMAZON

Deborah W. WOODCOCK*

ABSTRACT — Estimates of terrestrial biomass depend critically on reliable information about the specific gravity of the wood of forest trees. The study reported on here was carried out in the southern Peruvian Amazon and involved collection of wood samples from trees (70 spp.) in intact forest stands. Results demonstrate the high degree of variability in specific gravity (ovendry weight/green volume) in trees at single locations. Three forest types (swamp, high terrace forest with alluvial soil, and sandy-soil forest) had values close to the average reported for tropical forest woods (.69). Two early successional forest types, which make up as much as 12% of the total vegetated area in this part of the Amazon, had values significantly lower (.40). An increase in specific gravity with increasing age of the tree, which has been reported in some spe­ cies of tropical-forest woods, is seen in a positive relationship between specific gravity and di ameter for a species prevalent in one plot. Increases in specific gravity with tree and forest age may be significant in estimating changes in carbon stores over time.

Key-words: wood specific gravity, tropical forest biomass, Southern Peruvian Amazon

Densidade Específica da Madeira de Árvores e Tipos de Floresta no Sul da Amazônia Peruana

RESUMO — Estimativas de biomassa em ecossistemas terrestres dependem de informações confiáveis sobre a densidade da madeira das árvores. Neste estudo, realizado no sul da Amazônia peruana, foram coletadas amostras de madeira de árvores (70 spp.) em florestas intactas. Os resultados demonstram a grande variabilidade na densidade específica (peso seco / volume fresco) entre as árvores de um único sítio. Três tipos de floresta (baixio de terraço alto, floresta sobre terraço aluvial alto argiloso, e floresta de terra firme sobre solo arenoso) tiveram valores de densidade específica próximos à média reportada para madeiras de floresta tropical (0,69). Duas florestas em fase sucessional, que constituem até 12% da área vegetada nesta parte da Amazônia, tiveram valores significativamente menores (média de 0,40). Um incremento da densidade específica com a idade da árvore, reportada anteriormente para algumas espécies de árvores de floresta tropical, foi também encontrado para uma espécie avaliada neste estudo, com uma relação positiva entre sua densidade específica e seu diâmetro. Os aumentos de densidade específica com a idade, tanto das espécies como das florestas, podem ser importantes para estimativas de mudanças temporais nos estoques de carbono.

Palavras-chave: Densidade específica de madeira, biomassa de floresta tropical, Amazônia peruana

* Dept. of Geography - University of Hawaii - 445 Social Science - Honolulu HI 96822 USA woodcock@hawaii. edu

INTRODUCTION

Information on the wood specific gravity of forest trees, an important factor influencing the amount of for­est biomass, is available in various databases around the world (Detienne

& Chanson, 1996). Such sources, however, are not complete and include only a portion of the diversity of tree species that occur in the tropics (Fearnside, 1997). Field-based re­search is an additional source of infor­mation about this important quantity.

This study reports values for wood spe­cific gravity for a variety of forest types in the southern Peruvian Amazon.

METHODS

The field sites were located in the Tambopata-Candamo Reserve in the southern Peruvian Amazon. This area of the Amazon has very high biodiversity (574 bird species within walking distance of the lodge where I was based), some portion of which is undoubtedly attributable to the diver­sity of forest types present. The veg­etation types sampled are as follows (Fig. 1)

1) An area of low, early-succes-sional vegetation growing alongside the Rio La Torre on a sandbar deposit. Dominant taxa are Cecropia and Salix spp. and other taxa typical of dis­turbed riverine habitats, [n = 32 (8 spp.), mean dbh = 10 cm]

2) Mature floodplain forest growing along the La Torre on alluvial soils within a meander bend. The smooth-barked Capirona is a distinc­tive element of these gallery forests. There is standing water at some times of the year and severe flooding occurs at a recurrence interval estimated at 10-12 years. A plot for long-term eco­logical monitoring is near the transect (Phillips & Gentry, 1994; Phillips, 1999). [n = 34 (22 spp.), mean dbh = 22 cm]

3) Clay-soil forest on an "upper terrace." This surface is actually an ancient floodplain of the Tambopata with an age estimated at 40,000 or more years (Salo & Kalliola, 1990). Pseudolmedia laevis and P.

macrophylla are dominant species at the site, [n = 36 (21 spp.), mean dbh = 23 cm]

4) Sandy-soil forest. Away from the main river system, blackwater riv­ers drain substrates distinctly sandy in nature. The sandy-soil forest sampled here was growing alongside the Aguasnegras River approximately 6 km from the confluence of the Tambopata and the La Torre, [n = 29 (20 spp.), mean dbh = 24 cm]

5) Swamp forest on an upper ter­race. This floodplain feature may be a meander lake or an ancient channel. The palm Mauritia flexuosa is com­mon in these swampy areas. The trees sampled are adjacent to a permanent monitoring plot (Phillips & Gentry, 1994; Phillips, 1999). [8 spp. sampled, mean dbh = 29 cm]

Stand surveys were carried out by sampling on either side of a 60-m line transect. For the low vegetation along the riverbank, the transect was laid out parallel to the river's edge where all the trees present were small in diameter (<10 cm). In the higher-statured forests, sampling was limited to trees with diameters >10 cm.

Wood samples were obtained by harvesting a section of the trunk, in the case of the small-diameter riverine trees, or by means of a 12-mm incre­ment borer, in the case of the larger trees. In all cases, the material col­lected was at breast height and repre­sents one sample per tree. Cores were sealed and stored frozen until mea­surements could be made so that the wood remained hydrated and above the fiber saturation point. Leaves were

Figura 1. Location of plots: 1) low vegetation growing along a river bank; 2) mature floodplain forets; 3) high-terrace forest growing on rich alluvial soil; 4) sandy-soil forest; and 5) swamp forest.

also obtained and identifications are available for the majority of the trees sampled; voucher specimens are in the herbarium in the Department of For­est Sciences, National Agrarian Uni­versity-La Molina in Lima (MOL).

Specific gravity is determined here as oven-dry weight/green volume (with specimens oven-dried at 103° C for 24 hr). This measure represents density relative to water(density of 1 g.cnr 3) and thus is dimensionless. Val­ues are for the outer sapwood and were determined on pieces approxi­mately 2 cm long.

Altogether, 130 trees were sur­veyed and values of specific gravity were obtained for 70 species. Wood specific gravity for the various forest types is presented by species. Differ­ences between forest types were tested for by analysis of variance (Tukey-Kramer HSD test with the significance level set at .05). Values for specific gravity averaged by individual were also determined and represent either the average of all individuals present or a value that is weighted by number of individuals of each species present in the plot. Specific gravity for the

species studied are presented in the appendix.

RESULTS

Range of variation in specific gravity. Specific gravity is known to be a variable quantity (Detienne & Chanson, 1996), especially in tropical forests (Fearnside, 1997). Wood spe­cific gravity of the Peru woods (Fig. 2) varies by an order of magnitude, and variation of almost this degree is present at single locations.

An example of the degree of variation present is seen in plot 2, where the least dense wood is that of Erythrina ulei with value of .10 (close to that of Ochroma) and the most dense that of Minquartia guianensis with specific gravity of .70. The dif­ference between these woods is quite apparent under the microscope since the extent of stained (i.e., cell wall) material correlates well with specific gravity. The degree of variation exhib­ited is rather remarkable considering that both species reach the forest canopy and compete with one another for light and water. Clearly, different adaptive types are represented. The light-wooded Erythrina stores large amounts of water in its wood and dur­ing the short dry season, drops its leaves and is able to stay hydrated due to decreased evaporative demand. Minquartia, on the other hand, is ev­ergreen and continues to photosynthe-size during the dry season, aided by thick leaves resistant to desiccation. Other researchers have also related the range in specific gravity in tropical forests to niche part i t ioning

(Williamson, 1984; Borchert, 1994), with Borchert (1994) recognizing a variety of functional types in dry tropi­cal forest based on differences in spe­cific gravity and other aspects of tree biology.

Differences between vegetation types. The sample plots included two early successional associations grow­ing on low riverside terraces subject to frequent to occasional flooding. Trees in these plots have values of specific gravity (average of .40) significantly lower than those in mature clay-soil forest, swamp forest, or sandy soil for­est (Fig. 2). All of the latter have val­ues of specific gravity much closer to averages for tropical forest woods ( .65-.69; Brown & Lugo, 1992; Fearnside, 1997). In this part of the Amazon, as much as 12% of the for­ests may be actively disturbed by river processes (Salo & Kalliola, 1991) and thus in early stages of forest succes­sion characterized by trees with lower-specific-gravity wood.

Many pioneer species have light woods. Cot tonwood and aspen (Populus spp.) are temperate-latitude examples, although these woods, with specific gravity about .30, are not as light as the lightest tropical woods. It also appears to be generally the case that early successional forests, even in areas of high diversity like the study area, are characterized by trees with low specific gravity wood. Saldarriaga (1989) has reported similar results for areas recovering from clearing in the northern (Venezuelan) Amazon. That study showed that specific gravity av­eraged .54 in the 20 years after clear-

1.2

> .8

O) g .6 o CD C L W . 4

X I χ

1 2 3 4 5

site

Figura 2. Specific gravity by species for different vegetation types in Tambopata. 1) low river­side vegetation (8 spp., n= 32); 2) floodplain forest growing on a low terrace (22 spp., n= 34); 3) clay-soil forest on upper terrace (21 spp., n= 36); 4) sandy-soil forest (20 spp., n= 29); 5) swamp forest on upper terace (8 spp.). For each site, plots are as follows: (left) specific gravity by species with species representing >20% of the stand indicated by x's; (middle) box plots showing the median, the boundary between the middle and outer quartiles (ends of box), and range (bars); (right) averages by individual - either for all individuals or weighted by number of individuals of each species (stand data not available for plot 5).

ing, - .60 for years 30-80, and between .66 and .68 for mature forest.

A few caveats are in order in considering the significance of these results with respect to biomass esti­mates. First, the data presented here are for outer sapwood and may not be representative of average values. (Note, on the other hand, that pub­lished values for specific gravity are generally for heartwood.) Second, even though specific gravity averaged by individual (Fig. 2) also shows a difference between early and later-suc-cessional forests, it is not certain that these differences would show up in volumetric representations of forest biomass.

Variation within trees. The

prevalence of one species {Pseudolmedia laevis) in plot 3 pro­vided an opportunity to see how spe­cific gravity varied with diameter in this species (Fig. 3). The positive re­lationship found (correlation of .84) is understandable in terms of biome-chanical principles since structural re­inforcement in the outer part of the trunk is an economical way of achiev­ing strength and rigidity (Mosbrugger, 1990). In its early life, the tree puts its energy into growing up toward the light, only gradually producing wood of greater density at the periphery where it is more effective in support. The intensely competitive environ­ment of the forest may serve to inten­sify this growth pattern. Although

. 550 12.5

ι 1 1 1 1 15.0 17.5 20 .0 22.5 25.0 27.5

diameter (cm)

Figura 3. Specific gravity vs. tree radius for Pseudolmedia laevis. r= .84 (p <.05)

Wiemann & Williamson (1989a) re­port that pioneer species with low-den­sity wood are most likely to display such increases, P. laevis has specific gravity near the average for tropical forest woods.

Information on the directional trends most typical of tropical forest trees is incomplete. Increases were found in >50% of species in a variety of forest types in Costa Rica (Wiemann & Williamson, 1989a;b). Generally the degree of increase in specific gravity is on the order of . 1 -.4 (radial increase from inside to out­side of tree; Wiemann & Williamson 1988; 1989a;Omolodun et al, 1991; Butterfield et al 1993; de Castro et al, 1993; Woodcock et al, 2000). The considerations pertaining to support noted above also suggest that increases in specific gravity are common if not prevalent.

Significance for biomass esti­

mates. Of the various techniques used to estimate forest biomass, only one involves direct weighing of trees, lit­ter, and soil in forest plots (Fearnside 1985; 1987). But because the sample plots are by necessity limited in size, the high degree of heterogeneity in species composition and structure of the world's tropical forests means that it is difficult to obtain representative samples using this method. Another technique is to base analysis on values of specific gravity that are averaged by forest type or region and forest char­acteristics such as tree height and di­ameter that are either 1) known from forest inventory data or 2) can be de­termined in the field. Specific gravity is clearly a source of uncertainty for these determinat ions (Fearns ide , 1997).

Yet another approach to biomass estimation relies on studies in which trees of various sizes and representing

different species have been assessed by direct weighing. Generalized rela­tionships between biomass and tree characteristics included in forest in­ventories (primarily diameter) are then derived and applied to inventory sta­tistics to to arrive at biomass esti­mates. Since the forest inventory data collected and maintained by govern­ments around the world have broad spatial coverage, it has been possible to estimate forest biomass of large ar­eas such as the Brazilian Amazon and the eastern US in this way (Brown & Lugo, 1992; S. Brown et al, 1997; Schroeder et al, 1997). But although variation in specific gravity with age/ size of trees may be incorporated into the derived relationships, one equation is used to represent large biotic regions (Amazon forest, Eastern Deciduous Forest). If, as suggested here, there can exist significant differences in specific gravity within regions, these differ­ences could quite likely affect biom­ass estimates.

There exists considerable evi­dence of specific gravity increases with age/diameter in individual tropi­cal forest trees, although how wide­spread this pattern is, and the extent to which this variability is a source of error in biomass estimates, remains a question. It is notable that the trend is the same as that for forests — that is, specific gravity appears to increase both with increasing age of trees and increasing age/successional status of forests. The implication is that older forests are proportionately more im­portant biomass sinks than younger

forests for reasons additional to the presence of large trees and the in­creased carbon present in woody de­bris, leaf litter, and soils.

ACKNOWLEDGEMENTS

The author appreciates the assis­tance of Carlos Reynel, conversations with research staff at the Woods Hole Research Center, comments from Roger Hernandez, and support from NSF Climate Dynamics program (ATM 97-07899).

Literature cited

Borchert, R. 1994. Soil and stem water stor­age determines phenology and distribu­tion of tropical dry forest trees. Ecology, 75:1437-1449.

Brown, I. F., Martinelli, L.A.; Thomas, W.W.; Moreira , Μ.Ζ. ; Ferrei ra , C.A.C. ; Victoria, R.A. 1995. Uncertainty in the biomass of Amazonian forests: An ex­ample from Rondônia, Brazil. Forest Ecology and Management, 75:175-189.

Brown, S.; Lugo, A.E. 1992. Aboveground bio­mass estimates for tropical moist forests of the Brazilian Amazon. Interciencia, 17:8-18

Brown, S., Schroeder, P.; Birdsey, R. 1997. Aboveground biomass distribution of US eastern hardwood forests and the use of large trees as an indicator of forest development. Forest Ecology and Man­agement, 96: 37-47.

Butterfield R.P.; Crook, R.P.; Adams, R.; Mor­ris, R. 1993. Radial variation in wood specific gravity, fibre length, and vessel area for two Central American hard­woods: Hyeronima alchomeoides and Vochysia guatemalensis. International Association of Wood Anatomists Journal, 14: 153-161.

de Castro, F ; Williamson, G.B.; de Jesus, R.M. 1993. Radial variation in the wood spe-

cific gravity of Joannesia princeps: The role of age and diameter. Biotropica, 25: 176-182.

Detienne, P.; Chanson, B. 1996. L'Eventail de la densité du bois des feuillus. Bois et Forets des Tropiques, no. 250: 19-30

Fearnside, P.M. 1985. Brazil's Amazon forests and the global carbon problem. Interciencia, 10: 179-186.

Fearnside, PM. 1987. Summary of progress in quantifying the potential contribution of Amazon deforestation to the global car­bon problem. In: Athie, D.; Lovejoy, T.E.; Oyens, P.M. (Eds.). Proceedings of the Workshop on Biogeochemistry of the Tropical Rain Forests: Problems for Re­search. Universidade de São Paulo: 75-82

Fearnside, P.M. 1997. Wood density for esti­mating forest biomass in Brazilian Amazonia. Forest Ecology and Manage­ment, 90: 59-87.

Mosbruger, V. 1990. The Tree Habit in Land Plants. Springer-Verlag, Berlin. 161 p.

Omolodun, 0 . ; Cutter, B.E.; Krause, G.F.; McGinnes, E.A., Jr. 1991. Wood quality in Hildegardia barteri (Mast.) Kossern - An African pioneer species. Wood and Fiber Science, 23: 419-435.

Phillips, O. 1999. Long-term environmental change in tropical forests: Increasing tree turnover. Environmental Conservation, 23: 235-248.

Phillips, O.L.; Gentry, A.H. 1994. Increasing turnover through time in tropical forests. Science, 263: 954-958.

Saldarriaga, J.G. 1987. Recovery following shifting cultivation: A century of succes­sion in the Upper Rio Negro. In: Jordan, D. F. (Ed.). Amazonian rain forests: eco­system disturbance and recovery. Springer-Verlag, New York: 24-33

Saio, J.S.; Kalliola, R.J. 1991. River dynam­ics and natural forest regeneration in the Peruvian Amazon. In: Gomez-Pompa, A. et al. (Eds.). Rain Forest Regeneration and Management. Man in the Biosphere vol. 16, UNESCO and Cambridge Univ. Press, Paris and Cambridge: 245-256

Schroeder, P.; Brown, S.; Mo, J.; Birdsey, R.; Cieszewski, C. 1997. Biomass estima­tion for temperate broadleaf forests of the United States using inventory data. Forest Science, 43: 424-434.

Wiemann, M.C.; Williamson, G.B. 1988. Ex­treme radial changes in wood specific gravity in some tropical pioneers. Wood and Fiber Science, 20: 344-349.

Wiemann, M.C.; Williamson, G.B. 1989a. Ra­dial gradients in the specific gravity of wood in some tropical and temperate trees. Forest Science, 35: 197-210.

Wiemann, M.C.; Williamson, G.B. 1989b. Wood specific gravity gradients in tropi­cal dry and montane rain forest trees. American Journal of Botany, 76: 924-928.

Williamson, G.B. 1984. Gradients in wood specific gravity of trees. Bulletin of the Torry Botanical Club, 111:51 -55.

Woodcock, D. W.; Dos Santos, G.; Taylor, D. 2000. The Buttressed Blue Marble Tree: wood and growth characteristics of Elaeocatpus angustifolius (Elaeocarpaceae). Annals of Botany, 85: 1-6.

Aceito para publicação em 13/09/2000

Appendix. Specific gravity of Tambopata Woods. Asterisks indicates taxa constituting >20% of stand diversity.

Site Species Family Specific gravity (g.crrv3)

Cecropia ficilolia Warburg ex Snethlage Cecropiaceae

Enterolobium schomburgkb (Bentham) Bentham [1,2] Leguminosae

Ficus insipida Willdenow ssp. insipida [1,2] Moraceae

Margaritaria nobilis L. f. Euphorbiaceae

Miconia calvescens DC. Melastomaceae

Salix humboldtiana Willdenow Salicaceae

Sapium glandulosum (L.) Morong Euphorbiaceae

Visma angusta Miquel Gulti ferae

2 Erythrina ul& Harms Leguminosae

2 Eugenia tlorida DC. Myrtaceae

2 Ficus mathensii (Miquel) Miquel Moraceae

2 Guatteria olivacea R. E. Fries cf. Annonaceae

2 Inga semialata (Veil. Cone.) C. Martius [2,4] Leguminosae

2 Inga sp. Leguminosae

2 Minquartia guianensis Aublet Olacaceae

2 Ocotea sp. Lauraceae

2 Perebea anguslüolia (Poeppig & Endicher) C. C . Berg Moraceae

2 Pourouma cecropiifolia C. Martius [2,3] Cecropiaceae

2 Pourouma guianensis Aublet spp. guianensis Cecropiaceae

2 Sylogene cauliflora (Miguel & C. Martius) Mez Myrsinaceae

2 Symphonia globulifera L. f. Gutti ferae

2 Tabernaemontana flavicans Willdenow ex Roemer & Shultes Apocynaceae

2 ?Tetragastris Burseraceae

2 Theobroma cacao spp. sphaerocarpum (A. Chevalier) Cuatrecasas Sterculiaceae

.266

.399-.501

.281-.397

.557

.4

.343

.432

.488

.111

.634

.379

.387

.446-.449

.471

.754

.282

.517

.373-.561

.370

.535

.527

.487

.469

.430

2 Trichilia quadrijuga Η, & B. spp. quadrijuga Meliaceae .338

2 Unonopsis veneliciomm (C. Martius) R. E. Fries Annonaceae .436

2 Virola calophyllum Warburg Myristicaceae .342-.449

2 Xylopia cuspidala Diels Annonaceae .510

3 Casearia javitensis Η. Β. K. Fiacourtiaceae .736

3 Cecropia sciadophylla C. Martius Cecropiaceae .474

3 Ceiba pentandra (L.) Gaertner cf. Bombacaceae .489

3 Endlicheria bracteata Mez Luraceaea .573

3 Guarea glabra M. Vahl Meliaceae .592

3 Hevea guianensis Aublet [3,4] Moraceae .595-.609

3 Jacaranda copaia (Aublet) D. Don Bignoniaceae .402

3 Licania britteniana Fitsch [3,4] Chrysobalanaceae .675-.779

3 Naucleopsis ternstroemiiflora (Mildbraed) C. C. Berg Moraceae .612

3 Nectandra lucida Nees Lauraceae .682

3 Pourouma minor Benoist [3,4] Cecropiaceae .428-.481

3 Pouteria hispida Eyma Sapotaceae .758

3 Pouteria torta (C. Martius) Radlkofer Sapotaceae .629

*3 Pseudolmedia laevis (R & P.) J . F. Macbride Moraceae .563-.691

•3 Pseudolmedia macrophylla Trecul Moraceae .635-.686

3 Quiina florida Tulasne Guiinaceae .728

3 Quiina sp.? Quiinaceae? .489

3 Simarouba amara Aublet cf. Simaroubaceae .352

3 indet. Moraceae? .450

3 indet. Lauraceae? 1.022

4 Abarema jupunba (Willdenow) Britton & Killip Legiminosae .660

4 Calycophyllum spruceanum (Bentham) Hooker f. ex Schumann Rubiaceae .645

4 Casearia cf. decandra Jacquin Fiacourtiaceae .563

4 Conceveiba guianensis Aublet Euphorbiaceae .538

4 Corcfc'a scabrilolia D.C. Boraginaceae .474

4 Dialium guianensis (Aublet) Sandwith Leguminosae .481

4 Inga cf. chartacea Poeppig Leguminosae 481

4 Iryanthera juruensis Warburg [4,5] Myristicaceae .553-.685

4 Micropholis guyanensis (A. DC.) Pierre Sapotaceae .783

4 Nectandra cuspidata Nees Lauraceae .482

4 Stoanea Iragrans Rusby Elaeocarpaceae 4 7 0

4 Tachigali peruviana (Dwyer) Zarucchi & Herendeen Leguminosae 763

4 Wro/a sebifera Aublet Myristicaceae .500

4 indet. [pink wood - Aspidosperma?] Apocynaceae? .849

5 Brosimum lactescens (S. Moore) C. C. Berg Moraceae .804

5 Licaria armeniaca (Nees) Kostermans Lauraceae 553

5 Maquira coreacea (Karsten) C. C . Berg Moraceae .541

5 Nectandra sp. Lauraceae .342

5 Pseudolmedia sp Moraceae .787

5 indet Myristicaceae? .804

5 indet Leguminosae .787